Sensor to digital meter interface

[Quantised]

I have Lascar SP 300 digital meter (RS223-183) that normally accepts 200mv input or other inputs using scaling resistors.

The sensor has 16.7mv/psi output which I plan to buffer with op-amps and multiply by two so the full scale of the sensor output is 200mv.

The sensor and its drivers is 4 meters away from the meter and passes by some noisy equipment like alternators and generators and engines and motors etc so my first question is it OK to go with a 200mv signal maximum (typical would be 80mv) or will the signal be effected from local interference or should I raise the sensor output to say 2v full scale and then add the resistors to the meter to reduce the voltage back to 200mv?

The second question:- is it better to use the 0v as common for both sensor and meter or to use the InHgh and InLow pins.

Thanks

Q

 

[canadaelk]

Q1: Does your sensor have a balanced output?
Q2: Is your meter build around the 7106/7107 IC?
E

 

[ronv]

Sensor amp

Since you need to amplify the signal a little anyway, why not use an op amp capable of driving 50 ohm coax and terminate at the +/- input of the meter with 50 ohms. May need to adjust offsets on the amplifier due to the small signal and high current drive.

 

[JimB]

I have Lascar SP 300 digital meter (RS223-183) that normally accepts 200mv input or other inputs using scaling resistors.

The sensor has 16.7mv/psi output which I plan to buffer with op-amps and multiply by two so the full scale of the sensor output is 200mv.

The sensor and its drivers is 4 meters away from the meter and passes by some noisy equipment like alternators and generators and engines and motors etc so my first question is it OK to go with a 200mv signal maximum (typical would be 80mv) or will the signal be effected from local interference or should I raise the sensor output to say 2v full scale and then add the resistors to the meter to reduce the voltage back to 200mv?

I suggest that the signal wiring (and the power wiring) are:
a/ twisted pairs to minimise magnetic pick-up
b/ kept as far as possible from the electrically noisy bits

You could also use a screened cable for the power/signal from the sensor probably earthing the screen only at the meter end.

A coax cable terminated in 50 ohms? Err... no I dont think so!

If you are doing the signal processing at the sensor, then you could amplify to 2volts and attenuate at the meter, cant do any harm.
Or, if you do the signal processing at the meter, use a simple RC low pass filter to remove the noise.

The second question:- is it better to use the 0v as common for both sensor and meter or to use the InHgh and InLow pins.

I think this depends a lot on the sensor and how you intend to power it.

JimB

 

[Quantised]

canadaelk - the sensor is centered around 6v and is then referenced to 0v, the display is driven by a built in MAX140.

JimB - The conditioning is done at the sensor end so it has voltage followers close to the sensor and an op amp to power the signal up the line to the meter.

It seems a good idea to run it at a higher voltage then divide it and also put a RC filter at the meter end, the fc could be set at around 1 or 2 hz maybe?

The preliminary design is attached.

Thanks

Q

 

[ronv]

Shielded twisted pair

I think this is the best for twisted pair. Your 6462 is pretty low power. Is this like a dc sensor?

 

[Quantised]

The sensor is a RS 286 664 and is a piezo resistive bridge driven from +12v to 0v and the output floats around 6v nominal.

 

[Reloadron]

I may be wrong here but I see a few things differently. First the output of your bridge is 100 mV Full Scale for an input of 15 PSIG. That works out to a sensitivity of 6.666 mV per PSI or roughly a sensitivity of 6.67 mV / PSI. That is from the specification sheet and based on 10 VDC excitation to your bridge. The drawing you posted reflects a 12 VDC excitation which changes things.

As to wire? I would use Alpha manufacture part number 2414C or equivalent. That is foil shielded wire and will work fine. There is no need to amplify your signal.

The LASCAR little DPM will work fine. What you would normally have is a FS output from your transducer is 15 PSIG = 100 mV so your scaling resistors for the meter should be Ra = 910 KΩ and Rb = 150 KΩ. With the decimal set on the meter the meter will read 15.0 with an input of 100 mV.

Again, if you choose to use 12 Volt excitation to the bridge the FS output of the bridge will not be 100 mV. The FS output would then be 120 mV and Ra and Rb need changed. Ra would be 910 KΩ and Rb changes to 130 KΩ.

That is how I see it anyway.

Ron

 

[Quantised]

Ron Thanks for your input - oooops it was a typo in the 1st post and it is 6.67mv per psi as you point out.

If I use the scaling resistors then my input at FS of 100mv will be 15mv so I am only using 15mv out of a range of 200mv of the meter. This is why I thought to multiply by 2 to get 200mv full scale from the sensor and 150mv would represent 15psi on the meter. Set the DP to show 15.0 - this way it uses more of the meters range and so avoids errors being an issue - well that was the idea anyway.

I like the shielded wire idea and already have a 5 core signal with braided wire shield in situ from an old and now thrown instrument - I could use that?

Also I read with great interest the excitation voltage, this is new to me and I had never thought of it but realise now I should have. It is even more important as the voltage is a nominal 12v (engine electrics) so could vary from 12v up to 14.4v as charging and loads occur. Thats bad for calibration so I shall plan to use a BZX85C10 with a R470 from Vs to stabilise the votage to the sensor

The sensor has a 5K resistance from Vs to 0v - just measured it.

Thanks Ron

 

[Reloadron]

Yes, before I forget the excitation voltage really needs to be a very stable voltage source.

When I am in a situation that requires using a digital panel meter to display engineering units I try to opt for a meter like this one. The downside to using a meter like this is obviously the cost factor. The plus side is in the case of the linked to meter is user scaling for engineering units and they offer an excitation output for the sensor being used. Additionally they can be powered from various voltages including a 9 to 32 VDC source.

Actually the linked to meter is on the low end of the price range for meters like this. But they do work extremely well.

Now in your case I would first try things as I originally suggested and not worry about amplification. Then if you feel it is needed advance to adding a small instrumentation amplifier stage. When done like this you actually want your amplification as close to the sensor as possible so you don't amplify noise. Rember, the excitation voltage needs to be stable, like a rock.

Ron

 

[Quantised]

OK I will make the excitation solid - good idea.

The meter you posted is interesting but I already have mine and the reason to use them is they are IP67 sealed so it can be used in harsh environments plus its small enough to fit the limited space on the panel.

I will go with the conditioning at day one as it will allow me to alter the sensitivity.

Great input - Thanks

 

[Reloadron]

Let us know how it works out.

Ron